Abstracts
predictors (p!0.001, 95% CI, 0.58-2.32) on multinominal logistic regression analysis. Conclusions: EUS-TCB is a safe with a diagnostic accuracy of 75% for tissue sampling. The diagnostic yield is highest if the target lesion is sampled via the transgastric route and more than two passes are made.
Pancreatic cyst sampling: to brush or not to brush? That is the question T. Thomas, K. Ragunath, P. V. Kaye, G. P. Aithal Background: Fluid analysis obtained by EUS guided FNA has been used to aid in diagnosis and management of cystic lesions in the pancreas. Complementing fluid aspiration with brushing of cyst wall may increase the cellular yield. Aim: To compare the cellular yield of pancreatic cyst wall brushings with that of fluid aspiration. Method: Patient with cystic pancreatic lesions referred for EUS evaluation and sampling as a part of their diagnostic work up between May 2002 and Jan 2008 were identified from an endoscopy database and demographic, clinical and cytological data collected. Patients were divided into brushings group (cyst aspiration and wall brushings) and fine needle aspiration group (cyst aspiration only). Samplings were performed using a Linear-array echoendoscope (GF-UCT240-AL5, Olympus, KeyMed), 19G/22G FNA needle and echobrush (Wilson Cook, Ireland). The cellularity was determined by a single histopathologist and divided into either adequate or scanty/acellular. Acellular yield and insufficient scanty yield were considered false negatives even if the final diagnosis was benign pathology. Results: 53 patients (Aspiration group: 27, Brushings group: 26,) with a median age of 69 years (IQR: 49-77) were included. There was no significant difference in age (p Z 0.34) cyst size (p Z 0.67) or cyst location (p Z 0.22) in both groups. Overall 24.5% (13/53) of samples were acellular/insufficient with no significant difference in both groups (aspiration vs. brushings, 22% vs. 27%, p Z 0.93) and the remaining samples adequate for cytological evaluation (aspiration vs. brushings, 77.7% vs. 73%). In total, 28 cases were neoplastic (18 benign, 10 malignant). The diagnostic accuracy was 62.9% v. 61.5% in brushings and aspiration groups respectively. 2/4 (50%) patients were diagnosed as cancer in the brushing group compared to 1/6 (16%) in the FNA group but this was not statistically significant (two sided p Z 0.40). 22/53 (41.5%) was true negatives (FNA only vs. brushings, 37% vs. 46%) at an overall median follow up of 10 months (12 and 10 months median follow up in both groups respectively). Conclusion: The cellular yield was similar in fine needle aspiration and Brushings group. Greater proportion of patients with malignant cystic pancreatic lesions diagnosed by EUS sampling was in the brushings group, but this did not reach statistical significance.
Analysis of K-ras point mutation in pancreatic mass lesions using endoscopic ultrasonography-guided fine needle aspiration K. Yamakita, A. Sawaki, N. Mizuno, K. Hara, T. Takagi, T. Nakamura, M. Tajika, H. Kawai, K. Yamao, K. Hosoda, Y. Yatabe Background: Differentiation between benign and malignant pancreatic mass lesions still remains a great challenge in clinical practice. Moreover, there are still some patients who undergo surgery and finally prove to have a benign disease. In this study, we investigated the diagnostic accuracy of EUSFNA and the frequency of Kras point mutation in the specimens obtained by EUS-FNA in various kinds of pancreatic mass lesions. Materials and Method: We reviewed 81 patients with pancreatic mass who underwent EUS-FNA from April 2005 to March 2008. The final diagnosis was made by surgery, EUS-FNA result or clinical follow up, more than 6 months. 81 patients included 63 pancreatic ductal carcinomas, 6 endocrine tumor, 10 tumor-forming pancreatitis(TFP), 1 solid pseudopapillary tumor (SPT) and 1 acinar cell carcinoma (ACC). The EUS-FNA specimen was also examined for the frequency of K-ras mutation. Result: The sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy of EUS-FNA for pancreatic malignancy were 88%, 100%, 100%, 67%, 90% respectively. The frequency of K-ras mutation was 84% in pancreatic ductal carcinomas. K-ras point mutation was observed in only one TFP case which was negative for cytology and re-examined 1.5 month later with negative K-ras mutation. K-ras point mutation was not observed in endocrine tumors, SPT, and ACC. There was significant difference between pancreatic ductal carcinomas and TFP or endocrine tumor in the occuerrence rate of K-ras mutation(p!0.05). When we defined the tumor with a positive K-ras mutation as a gemcitabine sensitive tumor(ductal carcinoma) of the pancreas, the sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy for gemcitabine sensitive tumor were 84%, 94%, 98%, 63%, 86% respectively. Conclusions: In most of the pancreatic mass lesion, gemcitabine sensitive tumor can be diagnosed by the combination of cytology and K-ras mutation. Histopathological and immunohistochemical studies should be performed only in selected cases of negative cytology or negative K-ras mutation.
www.giejournal.org
ABSTRACTS SELECTED FOR POSTER PRESENTATION The effectiveness of EUS in detecting non-insulinoma gastroenteropancreatic neuroendocrine tumours Yang Aiming, Shu Huijun, Qian Jiaming, Lu Xinghua Background: Endoscopic ultrasonography (EUS) is a recently introduced diagnostic technique for detecting abnormalities in gastroenteric walls, pancreas and adjacent structures. But the effectiveness of this technique in diagnosing some different diseases is not fully investigated. Objectives: To investigate the effectiveness of EUS in detecting non-insulinoma gastroenteropancreatic neuroendocrine tumors (GEPETs). Methods: Eleven patients with non-insulinoma GEPETs diagnosed by Peking Union Medical College Hospital between June 1996 and June 2007 were reviewed retrospectively. Echoendoscopic patterns of these patients were investigated and diagnostic sensitivity of EUS was compared with abdominal angiography and OctreoScan. Results: Eleven patients had altogether 16 primary tumors. The sensitivity of EUS in locating non-insulinoma GEPETs was 75.0% (12/16). Eliminating one patient who underwent gastrectomy before, which obviously changed local anatomic structures, the sensitivity of EUS was 90.9% (10/11). The sensitivities of abdominal CT, US, OctreoScan and angiography were 56.3% (9/16), 37.5% (6/16), 66.7% (8/12) and 57.1% (4/7) respectively. As to echoendoscopic patterns, of the 12 tumors identified by EUS, 8 were hypoechogenic, 2 were isoechogenic, and 2 were hyperechogenic; 5 were homogeneous and 7 were inhomogeneous; 11 were well delimited. The smallest detected case by EUS was 5mm in size. Conclusions: EUS is a sensitive technique in locating non-insulinoma GEPETs.
EUS features in patients with autoimmune pancreatitis Yang Aiming, Wu Xi, Qian Jiaming, Lu Xinghua, Wu Dongsheng, Yao Fang Objective: To evaluate the EUS features in the patients with autoimmune pancreatitis (AIP). Methods: Reviewed eleven patients with AIP retrospectively. Results: 81.82% (9/11) of the patients had obstructive jaundice and 63.64% (7/11) had weight loss. 5 patients had the history of diabetes, and 3 presented with symptoms related to extrapancreatic organ involvement, such as salivary gland and lacrimal gland. In addition to obstructive jaundice, most patients had hepatic cell enzymes increasing remarkably and erythrocyte sedimentation elevating. Some had positive serum antibodies. CT showed diffuse and focal enlargement of pancreas, minimal peripancreatic stranding and dilation of extrahepatic bile duct. Stenosis in the distal bile duct and pancreatic duct involvement were demonstrated in MRCP or ERCP. The most common finding on EUS is diffuse or focal enlargement of pancreas along with hypoechoic parenchyma and a wavy margin. In addition to dilation, the bile duct had a prominently thickened wall with homogenous mild hypoechoic image. Peripancreatic lymph nodes could be enlarged. No pancreatic duct dilation, calcification or cysts were found. No peripheral vessels involved. Six patients had cytologic results from fine needle aspiration (FNA) by EUS, and pancreatic carcinomas were excluded. Conclusion: EUS could demonstrate specific images which increase the possibility of AIP diagnosis.
EUS-guided celiac plexus neurolysis vs. block: assessing interference with pancreatic surgery and adhesions in a pig survival model Vasu Appalaneni, Katherine Morgan, Robert Hawes, Brenda Hoffman, Joseph Romagnuolo Background/Purpose: Celiac plexus neurolysis (CPN) is used for pain management in pancreatic cancer. It involves the injection of alcohol to destroy the celiac neural structures, aimed at achieving longer-lasting analgesia than a steroid/anesthetic block alone (CPB). However, one concern in chronic pancreatitis is that subsequent surgery may prove difficult due to adhesions. The aim of the study was to determine the degree of interference and the adhesion burden from CPN vs. CPB in a blinded pig survival model. Methods: A total of 12, 25kg male Yorkshire pigs were used for the study, undergoing survival EUS-guided celiac single injection under general anesthesia with one of three solutions. Group 1 (nZ5) underwent CPN using dehydrated alcohol after 20cc 0.25% bupivicaine. Group 2(nZ5) underwent CPB with 20 cc 0.25% bupivicaine and 80mg triamcinolone. Group 3(nZ2) had 20cc saline injected. Video recording to judge the ease of distal pancreatectomy was performed. The primary endpoint was assessed at non-survival distal pancreatectomy, 6-14 days after the celiac injection, by a pancreatic surgeon, who was blinded to the type of injection each pig recieved. Interference with surgery (including identification of the SMV-neck-of-pancreas plane, important in Whipple surgery) was assessed using the following scale: 0 (none), 1 (minor), 2 (mild), 3 (moderate), 4 (severe interference). The composite adhesion score (as used in
Volume 69, No. 2 : 2009 GASTROINTESTINAL ENDOSCOPY S233